The Effects of Carboxylic Acids in Aluminum Anodizing

University of New Hampshire University of New Hampshire Scholars' Repository

Honors Theses and Capstones Student Scholarship

Spring 2017

The effects of carboxylic acids in aluminum anodizing

Abby E. Koczera University of New Hampshire

Follow this and additional works at: https://scholars.unh.edu/honors

Part of the Metallurgy Commons, Other Chemical Engineering Commons, Other Materials Science and Engineering Commons, Semiconductor and Optical Materials Commons, and the Transport Phenomena Commons

Recommended Citation Koczera, Abby E., "The effects of carboxylic acids in aluminum anodizing" (2017). Honors Theses and Capstones. 330. https://scholars.unh.edu/honors/330

This Senior Honors Thesis is brought to you for free and open access by the Student Scholarship at University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Honors Theses and Capstones by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected]. The effects of carboxylic acids in thermodynamics depends upon the concentrations and standard reduction aluminum anodizing potentials of bath components, pH, and possible reactions. Kinetics describes rates Abby Koczera of reactions, and in this case for homogenous May 2017 Honors Thesis reactions in solution and heterogeneous Department of Chemical Engineering reactions at the electrode surface, including University of New Hampshire faradaic reactions that entail the transfer of electrons and the amount of applied potential. Durham, New Hampshire 03824 In anodizing, reactants transport to the

surface of the solid-liquid interface and Abstract products move away via diffusion and convection.1 Hard-anodized alumina coatings were Aluminum anodizing is an electrolytic formed in sulfuric acid at low temperature process that is used to coat the metal with a and high current density in the presence of protective oxide layer. The oxide coating is carboxylic acid additives. Citric acid, formed on the aluminum by passing an trimesic acid, mellitic acid and electrical current through an acidic anodizing ethylenediaminetetraacetic acid (EDTA) bath. The coating protects the aluminum were utilized in varying concentrations. The beneath it, resisting corrosion and abrasion additives were chosen for their capacity to much more efficiently than raw aluminum.2 form complexes with tri-valent aluminum The anodizing process includes the following and hence impart chemical stability to the reactions in which the aluminum metal is coatings. The coatings were sealed in boiling oxidized and further reacted with water to water, and corrosion resistance was observed form alumina. in a high pH solution of potassium hydroxide. The coatings were examined using scanning Al  Al3+ + 3e- (1) electron microscopy (SEM) to assess coating 2Al3+ + 3H O  Al O + 6H+ (2) thickness and pore dimensions. Thicker 2 2 3 coatings were produced when the additive Alumina is a very hard material at all pH inhibited oxide coating dissolution, values. It is also corrosion resistant, but only increasing corrosion resistance. Overall, at neutral pH. It is vulnerable to corrosion at carboxylic acid additives showed a positive high and low pH, such as household cleaning impact on corrosion resistance when coupled supplies and food product, respectively. with sealants. More research in this field Previous studies have been done to show how could improve products used in cleaning and lithium additives help make alumina more cooking environments to withstand resistant to corrosion. Although this research conditions of high and low pH. used one additive with the presence of lithium, the main objective was to address the Introduction benefits of adding organic carboxylate molecules to increase the stability and The basic principles of chemical complexity of the anodized coating structure. engineering are applied in this field of In anodizing, there are three main electrochemistry and include material and categories of additives: metal cations, energy balances, thermodynamics, kinetics, complexing organic compounds, and surface and transport. Material and energy balances active organic compounds. Although this incorporate heats of reaction and rates of research was focused primarily on the effect generation and consumption; of complexing organic compounds, a 1 background on metal cations may be carboxylic acid is essentially connected by a beneficial to understand the overall role of four-carbon chain. EDTA is known as a the specific additives used. Transition metal chelating agent that will “trap” trivalent cations are generally added in simple salt aluminum on contact. Corrosion tests in form to mitigate the effects of by-products; if basic solutions were performed to measure they exist in more than one valence state, they the effects of additives with different possess the potential to oxidize or reduce by- carboxylic acid concentrations. products in the bulk of the bath. Reduction (a) Citric Acid potentials must be sufficiently positive or negative to drive oxidation or reduction, respectively. This idea may be used to compare the properties of the coatings based on the additives used. It is predicted that low concentrations of additives provide superior surface finishes, process stability, and 1 (b) Trimesic Acid uniform current distribution.

- + (a) O2 + 4e + 4H  2H2O + - (b) H2SO4 + 2H + 2e  H2SO3 + H2O - 2- (c) 2H2SO4 + 2e  S2O6 + 2H2O Figure 1. Possible reactions and their standard reduction potentials: (a) 1.229V, (b) 0.172V, (c) 1 -0.22V.

The formation of aluminum complexes (c) Mellitic Acid with organic molecules is one of the most common additive mechanisms. In this case, the reaction of hard-ion carboxylates with trivalent aluminum cations will readily form complexes that result in insoluble metal soaps that are incorporated onto the surface of the anodic coating. Because the pKa value of carboxylic acids is higher than the pH of sulfuric acid baths, the molecules are expected to protonate and become neutral in (d) EDTA solution, lacking the tendency to migrate toward the anode. Complexing additives form a thin film on the oxide surface to promote protection of the metal.1 Additives that were used in this study include citric acid, trimesic acid, mellitic acid, and EDTA, the structures of which can be seen in Figure 2. Citric acid contains three carboxyl groups, connected by a five-carbon chain. Trimesic acid contains three carboxyl groups around every other carbon on a benzene ring; Mellitic acid contains six carboxyl groups Figure 2. Chemical structures of each additive.3 branching off of a benzene ring; each 2

hydroxide) were placed in the chilled sulfuric acid solution, held by clamps and suspended Experimental above the bottom of the beaker. Different Aluminum Samples additives were added by mass (10mg, 100mg) to the solution at room temperature Aluminum 6063 tubes, produced by K&S and the process repeated. Precision Metals, were utilized for the Wires from the galvanostat connected the experiments. The rods (0.25 inch diameter) circuit, as seen in Figure 3. Once the were cut to a length of about six inches, ends equipment was turned on, the cell was left crimped, and covered with shrink wrap to untouched. PowerSuite software was utilized expose a controlled area of four square to measure potential versus time during centimeters for anodizing. The rods were left chronopotentiometric hard anodizing, with exposed at the top half in order to make an initial current step of -140.0 mA for 0.100 contact with the working electrode. seconds. Each experiment was run for 40 2 Chemical Solutions minutes with a current density of 35 mA/dm , and potential was measured with a voltmeter The 10% volume anodizing bath solution every two minutes. Once complete, the was created with ACS grade J. T. Baker samples were rinsed with deionized water, sulfuric acid (96.4%). 200mL of this solution dried, and placed in a labeled envelop for was placed into a beaker and further in an ice treatment and testing. The anodizing bath bath, left to chill for an hour prior to was saved and utilized about three to four anodizing. Different additives that were times before discarding. added to the solution include citric acid, trimesic acid, mellitic acid, and EDTA Treatments (Sigma-Aldrich, 99%). The collection of samples anodized with Solutions of potassium hydroxide were various additive concentrations were treated created for cleaning the rods (2M) and to to test the effects on corrosion resistance. perform corrosion tests (0.5M) using ACS The rods were treated in boiling water for grade pellets produced by Fisher Scientific fifteen minutes. (88.4%). Boiling water was used for coating treatment prior to corrosion tests. Anodizing The general procedure for anodizing began by cooling sulfuric acid solution in an ice bath for one hour (200mL, 10% volume). During that time, an aluminum sample was cleaned in potassium hydroxide (10mL, 2M) until the protective manufacturer coating was stripped off and clean aluminum metal was visible over the entire working area. Once removed from the basic wash, the sample was thoroughly rinsed with deionized water and placed in the electrolyte cell as the working electrode. The working electrode and counter electrode (titanium rod, cleaned in potassium

dried, and weighed; total mass lost was calculated. Scanning Electron Microscope (SEM) A selection of samples were viewed under the SEM to observe structural variations in the anodized coatings based on the presence of different additives. The rods were cut radially to view the coating thickness and pore behavior. Pictures were taken at various magnifications, with a maximum working distance of 9.71 mm, a beam intensity of about 5, and an applied voltage of 6.0kV. Coating thickness and pore dimensions were measured using Tescan analysis software. Image quality, judged by the appearance of streaks or uneven shading, is a result of the microscope resolution and the ability of electrons to reach the detector. Results and Discussion The following figures are a compilation of the data obtained from numerous trials of Figure 3. A schematic of the cell. anodizing. Potential is the driving force of the reaction. The steeper slopes indicate Corrosion Tests higher potential increase with time and thus a thicker coating, whereas the more gradual Corrosion tests were completed in caustic slopes indicate a thinner coating in the solution. For each test, the initial masses of presence of an additive with oxide coating three anodized samples were recorded using dissolution properties. a standard scale. The samples were then Coating thickness can be related to the placed in approximately 150mL of KOH behavior quantized in the graph in Figure 5. (0.5M). Times were recorded at which Higher potential increases with time indicates noticeable behaviors were seen, including the the growth of a thicker coating. The addition appearance of bubbles and their progression, of additives produced a thinner coating, with the appearance of streaks on the metal, the the exception of 10mg citric acid, which had presence of clean metal and how long it took a higher potential increase than the control for the entire anodized area to be corroded, sample. It is anticipated that thicker coatings etcetera. The way the coating was stripped have a higher resistance to corrosion. from the rod was observed as well, Trimesic acid samples differed little from sometimes falling off in multiple milliliter- the control sample. Although very similar, long delicate flakes, or polluting the solution the higher additive concentration had a in the shape of particles smaller than a grain slightly lower potential increase than the of sand. After a period of 60 minutes, the lower additive concentration. samples were wiped of residual smut. The Mellitic acid showed the same potential corroded samples were rinsed with water, behavior as the control sample. It was slightly lower than trimesic acid overall, so 4 this implies that mellitic acid will have measurements of each coating; contrasting thinner coatings of the pair. Although there trends may be due to experimental error. is virtually no distinction between the concentrations of mellitic acid, the higher sits below the lower, similar to the trend with trimesic acid. Of the entire spread, EDTA potential increase sits the greatest magnitude below the control sample. EDTA may have allowed or enhanced the dissolutive properties of the electrolyte, forcing a thinner oxide layer to result. After 30 minutes the concentrations diverge, contrary to the behaviors of the rest of the collection. The coating is created from the surface of the metal. As the coating gets thicker, the current has to pass through the growing layer to reach the clean metal surface. This mechanism explains that corrosion resistance improves as coating thickness increases. As the coating grows from the bottom up, it may simultaneously be dissolved at the surface. Figure 5. Potential over time for each additive. The aggressiveness of the electrolyte is Dashed black line is the control sample (no dependent on the properties of the additives additives), light blue is 10mg and dark blue is in the acid. Thicker coatings are produced 100mg. when the additive inhibits oxide coating dissolution. Table 1 organizes the thickness

Table 1. SEM measurements and mass lost during corrosion testing for each sample.

Quantity Coating Thickness Pore Diameter Pore Spacing Additive Mass Lost (g) (mg) (μm) (nm) (nm) None - 46 23 68 0.06 Citric Acid 10 48 29 78 0.07 Citric Acid 50 41 21 73 0.02 Trimesic Acid 10 55 20 61 0.02 Trimesic Acid 50 46 23 69 0.02 Mellitic Acid 10 49 24 83 0.06 Mellitic Acid 50 50 25 80 0.06 EDTA 10 45 30 87 0.06 EDTA 50 42 31 87 0.06

The coating thickness for the control sample concentrations of additives result in thicker was measured to be 46 microns. Overall, coatings. Mellitic acid has a contrasting citric acid, EDTA, and trimesic acid followed trend, with the 50mg sample having a bigger the same trend where 10mg yielded a thicker thickness than the 10mg sample (Fig. 6). It is coating than the 50mg. The magnitude of the proposed that additional additive contributes difference varies, with trimesic acid having to coating dissolution, however this may not the biggest change and EDTA the smallest. be the effects of mellitic acid. This supports the hypothesis that lower

Figure 6. Average coating thickness measurements for each sample.

Figure 7 illustrates the relationship between pore diameter and consequently large pore diameter and pore spacing. Diameter is spacing. Both EDTA samples characterized on the ordinate (0-35nm), and pore spacing is this behavior, followed by mellitic acid with arrange by a color slide (60-88nm). It is the next largest characteristics. Overall, most expected that large pore diameters also have samples had very similar characteristics to large spacing between them. Small pore the control sample. EDTA and 10mg citric diameters and large spacing are intriguing acid had the biggest deviations, resulting in because that would imply a denser coating. wide pores and spacing. Tall, purple bars indicate a sample with large

Figure 7. Pore diameter (ordinate) and pore spacing (color scale) results for each sample using SEM measurements.

10mg EDTA 10mg Mellitic Acid

50mg EDTA No additives Figure 8. Cross-sectional SEM images of four anodized samples. Magnification is 2.77kx. Cross-sectional images were taken using a spacing were measured at a working distance SEM. Coating thickness was measured at a of 8.13 mm and a magnification of 369 kx. working distance of 5.11mm and a Figures 8-9 show a selection of images taken magnification of 2.77 kx. Pore diameter and of the samples.

Figure 9. SEM pore images of 50mg mellitic acid (left) and 10mg trimesic acid (right).

Corrosion Results After anodizing, each sample was sealed corrosion (Fig. 11). It is evident that citric for 15 minutes in boiling water and then acid added little corrosion resistance to the weighed. In groups of three, the samples were coating. placed in 300ml of 0.5M potassium The mass lost after corrosion testing hydroxide. Pictures were taken periodically should be proportional to the rate of throughout the duration of the corrosion test. corrosion. Additives make the solution The samples were observed over 60 minutes. more aggressive and may reduce the coating At full on corrosive attack, there were rapid thickness as a result. But, they do appear to bubbles coming from the surfaces of the impart corrosion resistance since there was coatings (Fig. 10). Aggressive bubbling set residual mass after the 60 minute testing on at 10 minutes for EDTA, 15 minutes for period. The additives probably form a citric acid, 20 minutes for trimesic acid, and protective, though thinner, coating. Of the 20 minutes for mellitic acid. It is possible that set, trimesic acid (50mg) and EDTA (10mg) trimesic acid and mellitic acid had the highest appeared to have some coating remaining resistance to corrosion. It was hard to after corrosion (Fig. 11). The rest of the distinguish resistivity between the samples of samples seem to have a clean finish from a different concentrations. After 60 minutes, completely removed coating. Figure 12 residual smut was wiped off and the sample depicts the relationship between coating was weighed to obtain mass lost. thickness and mass difference. Trimesic The mass lost after corrosion testing acid had the thickest coatings and lost the should be proportional to the thickness of the least amount of mass. Coupled by its coating. Residual mass after the 60 minute physical appearance (Fig. 11), it appears that testing period indicates a strong coating. Of trimesic acid may have had the best the set, trimesic acid (50mg) and EDTA resistance to corrosion. EDTA and mellitic (10mg) appeared to have some coating acid samples had the same mass differential remaining after corrosion (Fig. 11). The rest as the control sample. Citric acid at 10mg of the samples seem to have a clean finish had the highest loss of mass. Combined from a completely removed coating. Figure with a relatively thick coating, it is likely 12 depicts the relationship between coating that this concentration of citric acid was thickness and mass difference. Trimesic acid susceptible to corrosion. Citric acid at 50mg had the thickest coatings and lost the least had the lowest loss of mass and the lowest amount of mass. Coupled by its physical thickness. Citric acid samples appear very appearance (Fig. 11), it appears that trimesic clean post corrosion (Fig. 11). It is evident acid may have had the best resistance to that citric acid added little corrosion corrosion. EDTA and mellitic acid samples resistance to the coating. The results suggest had the same mass differential as the control an ideal concentration range for each sample. Citric acid at 10mg had the highest additive that is high enough to form a proper loss of mass. Combined with a relatively coating but not excessive to dissolve the thick coating, it is likely that this coating during anodizing. concentration of citric acid was susceptible to corrosion. Citric acid at 50mg had the lowest loss of mass and the lowest thickness. Citric acid samples appear very clean post

Citric Acid: 10 minutes Mellitic Acid: 10 minutes

Trimesic Acid: 5 minutes EDTA: 10 minutes Figure 10. Samples undergoing corrosion testing in 0.5M KOH. From left to right in each picture ranks additive concentration (0, 10, 50mg).

Figure 11. Samples after corrosion testing. The average mass lost was 0.05g, ranging from 0.02 to 0.07g. From left to right: no additives, 10mg mellitic acid, 50mg mellitic acid, 10mg trimesic acid, 50mg trimesic acid, 10mg citric acid, 50mg citric acid, 10mg EDTA, 50mg EDTA.

Figure 12. Comparison between mass lost and coating thickness for each sample. Tall, blue bars indicate resistance to corrosion.

Conclusion additive concentrations would improve corrosion resistance. This research explored the effects of The results showed that lower carboxylic acids on alumina coatings. The concentrations of additives inhibit coating three major steps of the project included dissolution, and additional additive anodizing, sealing treatments, and corrosion contributes to the dissolution. The results tests; a selection of the anodized samples were from pore diameter and pore spacing showed further analyzed using scanning electron little variation for trimesic, mellitic, and citric microscopy. During anodizing, carboxylic acid. 10mg citric acid and EDTA had the acid additives including citric acid, trimesic largest diameters and consequently, the acid, mellitic acid, and biggest spacing. Corrosion testing and mass ethylenediaminetetraacetic acid (EDTA) were differential indicated trimesic acid as the most added to the baths in various concentrations resistant sample. The results suggest an ideal and cell potential was measured over time. It concentration range for each additive that is was expected that a high potential growth rate high enough to form a proper coating but not correlates to thick coatings, and that lower

11 excessive to dissolve the coating during anodizing. Ways to improve these conclusions would References be to lengthen anodizing time and observe later behavior among the combinations, as (1) Barkey, D. A Chemical Process well as repeat measurements under SEM of Engineering View of Additives in samples under the same conditions to rectify Aluminum Anodizing, University of areas of experimental error. Overall, the New Hampshire Department of presence of carboxylic acids in alumina Chemical Engineering, (2012). positively impacted its corrosion resistance (2) Global Metal Finishing, About when coupled with sealing treatments, but Anodizing.http://www.globalmetalfin there was not a consistent pattern of trends that ishing.com/anodizing-faqs. (accessed could lead to a specific conclusion without Dec 28, 2015). further experimentation, data collection and (3) PubChem, U.S. National Library of evaluation. Medicine. Compounds: Glutaric Acid, Mellitic Acid, Lithium Citrate. http://pubchem.ncbi.nlm.nih.gov/comp ound/glutaric_acid. (accessed Dec 28, 2015). (4) Spathis, P. Effect of bicarboxylic acids on the protection of anodic coated Al- alloys against SCC, Proceedings of the Symposium on Advances in Corrosion Protection by Organic Coatings II, 95, 309-321, (1995).